Better prediction of overheating in new homes
A detailed study of three modern, energy-efficient flats has been carried out to improve the means of predicting indoor temperatures and the risk of overheating, when designing multi-residential buildings.
Overheating in modern homes – which are often designed with a focus on improving energy efficiency – is a growing problem and likely to be exacerbated by climate change. To counter this, it is important when designing buildings to reliably assess indoor temperatures and the potential for overheating. These are typically predicted with dynamic simulations, using Building Performance Simulation (BPS) tools.
BPS tools need accurate data on a complex range of issues in the areas of climate, site context, building fabric, building services and occupant behaviour. All of these bring high levels of uncertainty that make correctly predicting indoor temperature very difficult, and can lead to a gap between the expected and actual performance.
A BRE Trust supported PhD project has been conducted by Kostas Mourkos at Loughborough University, to improve BPS tools’ reliability when predicting overheating risks in homes in multi-residential buildings. This was achieved by studying in detail three modern energy-efficient flats located in London. The flats are representative of many high-density developments built in London in recent years.
Areas of overheating assessments that have been revealed as needing improvement by the analysis include:
- specifying input values for numerous parameters, such as the ventilation rates of a mechanical ventilation system,
- providing guidance on handling the thermal interaction between communal spaces and the assessed flat,
- examining different infiltration and exfiltration pathways.
- The analysis also identified the key parameters influencing the observed gap between predicted and monitored indoor air temperature. While demonstrating how such a gap can be efficiently bridged through Bayesian calibration, this research showed that predicting overheating accurately remains challenging.
The research recommended that an overheating assessment should incorporate sources of uncertainty (such as occupant behaviour), by providing a range of values – instead of a single value – of the desired Building Performance Indicator (BPI). It should also consider using less sensitive overheating metrics.
Kostas was supervised by Prof Christina Hopfe and Dr Rob McLeod at Graz University of Technology, Dr Chris Goodier at Loughborough University, and Dr Mick Swainson at BRE. For more information contact Kostas (email@example.com) or access the paper.
Figure 1 Thermal model of one of the case study flats in SketchUp.
Figure 2 Minimum and maximum indoor air temperatures considering the inherent uncertainty in the thermal model (left image); overheating prediction including the identified uncertainty in the analysis (right image).
Figure 3 Comparison between monitored indoor air temperatures and predictions from the pre-calibrated model and the calibrated model.